Structural equation modeling of the relationships between pesticide poisoning, depressive symptoms and safety behaviors among Colorado farm residents

Purpose: To use structural equation modeling (SEM) to test the theory that a past pesticide poisoning may act as a mediator in the relationship between depression and safety practices. Depression has been associated with pesticide poisoning and was more strongly associated with safety behaviors than workload, social support or health status of farm residents in a previously published report. Methods: A cross-sectional survey of farmers and their spouses was conducted in eight counties in northeastern Colorado. Depressive symptoms were assessed using the Center for Epidemiologic Studies-Depression (CES-D) scale. Exploratory and confirmatory factor analyses were used to identify symptoms most correlated with risk factors for depression and safety practices. SEM was used to examine theoretical causal models of the relationship between depression and poor health, financial difficulties, a history of pesticide poisoning, and safety practices. Results: Exploratory factor analysis identified three factors in the CES-D scale. The SEM showed that poor health, financial difficulties and a history of pesticide poisoning significantly explained the depressive symptoms. Models with an excellent fit for the safety behaviors resulted when modeling the probability that the pesticide poisoning preceded depression, but no fit was possible when reversing the direction and modeling depression preceding pesticide poisoning. Conclusions: Specific depressive symptoms appeared to be significantly associated with primarily animal handling and farm machinery. The order of events, based on SEM results, was a pesticide poisoning preceding depressed mood in relation to safety behaviors.